Stabilization of the craniocervical junction after an internal dislocation injury: an in vitro study

The Spine Journal 15 (2015) 1070–1076

Basic Science

Stabilization of the craniocervical junction after an internal dislocation injury: an in vitro study Kris E. Radcliff, MDa, Mir M. Hussain, BSb,*, Mark Moldavsky, MSb, Noelle F. Klocke, MSb, Alexander Vaccaro, MDc, Todd J. Albert, MDc, Saif Khalil, PhDb, Brandon S. Bucklen, PhDb a

Rothman Institute, 2500 English Creek Ave., Building 1300, Egg Harbor Township, NJ 08234, USA b MERC, Globus Medical, Inc., 2560 General Armistead Ave., Audubon, PA 19403, USA c Rothman Institute, 925 Chestnut St, 5th Floor, Philadelphia, PA 19107, USA Received 8 September 2014; revised 22 December 2014; accepted 1 February 2015

Abstract

BACKGROUND CONTEXT: Reconstructive surgeries at the occipitocervical (OC) junction have been studied in treating degenerative conditions. There is a paucity of data for optimal fixation for a traumatically unstable OC joint. In clinical OC dislocations, segmental fixation may be impossible because of vertebral artery injury or fracture. Segmental fixation of the occiput, C1, and C2 demonstrated maximum biomechanical stability in fixation of an unstable craniocervical dislocation. A biomechanical study comparing various points of cervical posterior screw fixation after recreating traumatic injury would illuminate relative advantages between the various techniques. PURPOSE: To determine the rigidity lost, if any, of segmental C0–C2 posterior screw fixation versus fixation skipping C1 at the OC junction, with or without a cross-connector. STUDY DESIGN: This study is a cadaveric biomechanical investigation. METHODS: Intervertebral motions and translations were recorded in seven specimens under conditions in the following order: intact, OC dislocation model with complete disruption of the cruciate ligaments, alar ligaments, and occipitoatlantal/atlantoaxial capsules (injury), segmental posterior fixation (SPF) with posterior instrumentation (ELLIPSE; Globus Medical, Inc., Audubon, PA, USA) at occiput, C1, and C2 levels, endpoint fixation (EPF) with posterior instrumentation at occiput and C1 level skipping C1, and endpoint fixation with a cross-connector (EPFC). Motion was

FDA device/drug status: Not approved for this indication (Posterior cervical screws at C1 and C2); Approved (Occiptal Plating System). Author disclosures: KER: Other: Globus Medical (biomechanical testing) (0); Other: Globus Medical (Salary); Royalties: Globus Medical (0); Consulting: Globus Medical (C), Depuy (B); Trips/Travel: Globus Medical (B), Stryker (B), Medtronic (B), Relievant (B); Grants: Depuy (C), Medtronic (B), Paradigm (B). MMH: Other: Globus Medical (Paid Employee [E per year]). MM: Other: Globus Medical (Paid Employee [E per year]). NFK: Other: Global Medical, Inc. (Salaried Employee, E per year); Stock Ownership: Global Medical, Inc. (750 stock options currently endowed but not held); Research Support (Investigator Salary): Globus Medical, Inc. (Salaried Employee, Staff/Materials): Globus Medical, Inc. (Salaried Employee, E per year; travel to conferences for other Global Medical sponsored studies also covered). AV: Royalties: DePuy (B-C), Medtronic (F), Stryker Spine (F), Biomet Spine (F), Globus (F), Aesculap (0-B), Nuvasive (F); Stock Ownership: Replication Medica (B), Globus (F), K-2 Medical (F), Paradigm Spine (F), Stout Medical (F), Spine Medica (C-D), Computational Biodynamics (B), Progressive Spinal Technologies (F), Spinology (B-C), Small Bone Innovations (D-E), Cross current (B-C), Syndicom (B), In Vivo (B), Flagship Surgical (C-D), Advanced Spinal Intellectual Properties (co-owner), Cytonics (B), Bonovo Orthopedics (D-E), Electrocore (C-D), Gamma Spine (B), Location based Intelligence (C-D), FlowPharma (B), R.S.I. (0-B), Rothman Institute and Related Properties (F), Innovative Surgical Design, Spinicity (C-D); Consulting: Stout Medical (F), Gerson Lehrman Group (B), Guidepoint Global (B), Medacorp (B), Innovative http://dx.doi.org/10.1016/j.spinee.2015.02.002 1529-9430/Ó 2015 Elsevier Inc. All rights reserved.

Surgical Design, Orthobullets; Board of Directors: AO Spine, Innovative Surgical Design, Association of Collaborative Spine Research, Spinicity; Grants: Styrker Spine, NuVasive, Cerapedics (F). TJA: Royalties: DePuy (F), Biomet (F); Stock Ownership: Paradigm Spine (C-D), ASIP (B), Biomerix (C-D), Breakaway Imaging (C-D), Crosstree Capital Partners (C-D), In Vivo Therapeutics (B), Invuity (C-D), Pioneer Surgical (C-D), Vertech (B), Gentis (C-D); Consulting: Facetlink (B-C), DePuy (B-C), United Healthcare (B), Biomet Spine (B-C); Speaking and/or Teaching Arrangements: DePuy Spine (B-C); Board of Directors: United Healthcare, Rothman Institute and Related Properties (F); Scientific Advisory Board/Other Office: Gentis (B), United Healthcare (B). TA is also affiliated with ASIP, PMIG, CSRS, SRS, IMAST, and AOA. SK: Research Support (Investigator Salary, Staff/Materials): Globus Medical Inc. BSB: Other: Globus Medical (Salary). The disclosure key can be found on the Table of Contents and at www. TheSpineJournalOnline.com. Authors MH, MM, NK, SK, and BB are paid employees of Globus Medical, Incorporated. Although no specific funding was provided, some authors are salaried employees of a medical device company in which there is a research budget, whereas other authors have royalty/consultant affiliations. * Corresponding author. Globus Medical, Inc., 2560 General Armistead Ave., Audubon, PA 19403, USA. Tel.: (1) 610-930-1800 ext 2727; fax: (1) 610-930-2042. E-mail address: [email protected] (M.M. Hussain)

K.E. Radcliff et al. / The Spine Journal 15 (2015) 1070–1076

1071

applied through a custom spine simulator with a pure moment load of 2.5 Nm and measured with motion capture markers attached to occiput (C0), anterior C1 ring, and C2. Flexion-extension (FE), lateral bending (LB), axial rotation (AR), and cranial-caudal (CC) motions were recorded in terms of C0–C2. Results were reported as a percentage of injured motion (injury5100%), unless otherwise stated. RESULTS: The injury significantly increased the motion to 165%, 263%, and 130%, during FE, LB, and AR, respectively, of intact. The CC translations increased to 164%, 254%, and 121% during FE, LB, AR, respectively, of intact. Segmental posterior fixation significantly reduced motion to 7%, 8%, and 1%, during FE, LB, and AR, respectively, of injury. Endpoint fixation significantly increased motion in FE, resulting in 12%, 6%, and 4% during FE, LB, and AR, respectively, of injury when compared with SPF. The EPFC construct decreased the motion compared with its counterpart to 8.6%, 5.7%, and 3.2% during FE, LB, and AR, respectively. CONCLUSIONS: All fixation constructs significantly reduced motion in all loading modes and CC translations, compared with intact and injury. The construct with the greatest stability against craniocervical dislocation included SPF with instrumentation at the occiput, C1, and C2. By skipping C1 using the EPF, FE and cephalad-caudal translations significantly increased compared with posterior fixation at every level. The addition of a cross-connector increased the stability but was not statistically significant. Ó 2015 Elsevier Inc. All rights reserved. Keywords:

Cervical screw fixation; Craniocervical; Dislocation; Subluxation; Biomechanics; Range of motion

Introduction The craniocervical junction (CCJ) comprises the occiput (C0), the atlas (C1), and the axis (C2). It is unlike any other region of the spine because of its unique anatomical shape and large range of motion (ROM) [1,2]. The unique configuration allows more flexion-extension (FE) and axial rotation (AR) than anywhere else within the cervical spine because of its uniquely shaped articulating joints [2–6]. In the absence of intervertebral discs, this large ROM in CCJ is primarily restricted by additional ligaments that are not present in other parts of the spine [1,4]. The cruciate and alar ligaments are two main ligamentous supports that surround the odontoid process (dens) of C2 in the CCJ. Although they are noted for their strength, they are susceptible to disruption from external forces [1,3,4,7,8]. Dislocation of the CCJ usually results from disruption of the internal craniocervical ligaments [9]. The occipitocervical (OC) joint capsules [9–11] and the cruciate ligaments [12–17] are the main stabilizers of the CCJ. There are specific subtypes of craniocervical dislocation injuries, including isolated atlantoaxial injuries or combined occipitoatlantaoaxial injuries [9]. Ligament disruptions are also one of the main causes for surgical intervention to improve stability at C1–C2 joint [3,11,16,18]. In craniocervical ligamentous dislocation injuries, OC plating with segmental cervical fixation is the preferred technique. Occipital cervical plating has been validated in degenerative models [6,11,15,19–22] and shown to result in a high fusion rate clinically [23]. However, there are limited studies describing the OC ligamentous dislocation injuries. The traumatic injury models from the literature included fractures [24]. A prior study of traumatic instability only included transverse and alar ligament sectioning and did not

include resection of the craniocervical joint capsules [19]. Occipitoatlantal joint capsule injury is an important component of craniocervical dislocation injuries and is necessary in a reconstruction model [9]. Although segmental fixation should provide superior biomechanical results, segmental fixation from C0–C2 is often challenging in the case of OC dislocation injuries because of other associated bony or vertebral artery injuries [10,25]. The purpose of this study was to test various types of craniocervical instrumentation in a specific craniocervical dislocation injury model. Materials and methods Specimens Seven fresh-frozen cadaver spines (C0–C3) were used in this study. The medical history of each donor was reviewed along with radiographs to exclude any specimens with spinal trauma, malignancy, deformity, or fractures that would otherwise affect the outcome of the test. The specimens were carefully dissected leaving ligaments, bones, and intervertebral discs of the cervical segments intact. The specimens were fixed at the occiput proximally with screws and at C3 distally using a 2:1 mixture of Bondo auto filler (Bondo MarHyde Corp., Atlanta, GA, USA) and fiberglass resin (Home-Solution All Purpose; Bondo MarHyde Corp). All specimens were double wrapped in plastic bags and stored at
20 C before testing. Test setup and data analysis Each specimen was thawed overnight and mounted on a custom 6 df spine motion simulator. A load-control protocol where an unconstrained pure bending moment of 2.5

1072

K.E. Radcliff et al. / The Spine Journal 15 (2015) 1070–1076

Nm was applied at a rate of 1.5 /s in FE, lateral bending (LB), and AR. Due to the viscoelastic properties of the spine, each loading mode was tested for a total of three loading cycles; only the final cycle’s data were analyzed. After testing, statistical analysis was performed using a one-way analysis of variance with repeated measures to determine the significant differences (p!.05) between experimental constructs. Testing conditions The following constructs were tested (Fig. 1): 1) Intact, to develop a baseline comparison (intact). 2) Removal of the clivus (which also disrupts the alar ligaments), sectioning of the transverse ligaments, and finally sectioning the occipitoatlantal and atlantoaxial articulating joints (injured). 3) Segmental posterior fixation (SPF) instrumented at occiput, C1, and C2 levels. 4) Endpoint fixation (EPF) with instrumentation at occiput and C2, skipping the C1 level. 5) Endpoint fixation with a cross-connector (EPFC) with instrumentation at occiput and C2, skipping C1 level. After intact motion testing (Fig. 1A), the second testing condition was injury state (Fig. 1B). This was considered to represent the disruption of all vertical stabilizing structures. For the purpose of this study, the vertical ligamentous structures (apical ligament, superior band of transverse ligament, alar ligaments, and tectorial membrane) extending from the clivus to C1 were considered a single unit. This is because of adjacent and confluent soft tissue structures making identification and isolation of individual ligaments difficult. The clivus was then completely resected, leaving only the internal craniocervical stabilizing ligamentous structures around the dens intact, resulting in uninhibited access to the transverse ligaments without disruption of the anterior longitudinal ligament, tectorial membrane, or C1 anterior ring. Next in the posterior view, a scalpel was passed through the C1 lateral mass to ensure complete resection of the transverse ligament; this created instability between C1 and C2 vertebral bodies. Finally, both C0–C1 and C1–C2 joint capsules were sectioned and exposed so that the occipital condyles and C1–C2 articulating surfaces were visually exposed. For C0–C1 exposure, the anterior occipitoatlantal joint capsules were sectioned by passing a scalpel horizontally from the lateral aspect of the odontoid to the lateral aspect of the occipital condyle. C1–C2 exposure was performed by passing the scalpel medially from the lateral aspect of the atlantoaxial joints toward the odontoid. Before injury creation, instrumentation was inserted using fluoroscopic imaging. All specimens were initially instrumented with SPF, consisting of an occipital plate at C0, lateral mass screws at C1, and pars screws at C2 (ELLIPSE

Occipito-Cervico-Thoracic Stabilization System; Globus Medical, Inc., Audubon, PA, USA; see Fig. 1C). Each specimen then had C1 lateral mass screws removed and the rods were replaced using EPF at C0 and C2 only (EPF) (Fig. 1D). Finally, a cross-connector was added between C1 and C2 (EPFC) (Fig. 1E). Instrumentation The occiput was instrumented with a midline plate and six bicortical screws (3.514 mm). C1 lateral mass screws were placed using a high-speed burr with the Harms and Melcher [26] technique [27]. A 30 mm partially threaded 3.5 mm screw was then bilaterally placed with a bicortical purchase at C1. At C2, interarticularis pars screws of 3.514 mm were placed. Measurements Rigid body markers were placed at C0, C1 anterior ring, and C2 vertebral body. From these markers, ROM was obtained using Optotrack Certus (NDI, Inc., Waterloo, Canada) Motion Analysis System. Range of motion was measured in FE, LB, and AR for each testing condition. Motion at the CCJ is primarily shear (translation) with angular rotation [14]. Additionally, dislocations inherently involve joint subluxation and/or cranial translation. Therefore, intervertebral translations in the cranial-caudal (CC) planes were also measured in each condition. Translation was defined as the change in position of one vertebral body compared with its adjacent vertebral body. Although CCJ contains two motion segments, in the present study, motion is discussed as a whole from C0–C2. Translation is reported as the cumulative subluxation of both joints. The average ROM and translation were calculated across all specimens and were compared with each other. Differences between instrumentation methods’ motions were calculated based on the injury’s ROM (injured state5100% unless otherwise noted). Although not included within the results, coupled motions were also recorded and can be seen in the figures of the Appendix. For example, when the moment was applied in FE, coupled motions in LB and AR were recorded. Additionally in the Appendix, anteroposterior, medial-lateral, and CC planes were measured and recorded in all three planes during the applied rotations. Lastly, the Appendix also presents data separated into each motion segment of C0–C1 and C1–C2.

Results The injury increased the ROM from intact in all loading modes (Table), but was only statistically significant in FE and LB (164.8%6255.8% and 263.4%6239.9%, respectively, compared with intact). The injury also increased AR to 130%6131.7% of intact; however, this was not statistically

K.E. Radcliff et al. / The Spine Journal 15 (2015) 1070–1076

1073

Fig. 1. Image of the testing constructs: (A) intact; (B) injured with zoomed in image of both joints subluxed; (C) segmental posterior fixation with screws at every level of the craniocervical junction; (D) endpoint fixation with C1 screw removed; and (E) endpoint fixation with a cross-connector between C1–C2.

significant to intact. The CC translation also increased significantly to 164.2%6109.8% and 254.56250.5% during FE and LB, respectively. The CC translation during AR had a negligible increase of 121.3%648.8% of intact. All instrumentation constructs were significantly (p!.05) less than intact and injured. Segmental posterior fixation significantly decreased ROM to 7.0%65.7%, 5.3%64.1%, and 3.0%62.5% in FE, LB, and AR, respectively, compared with injury. Similarly, the EPF construct significantly reduced ROM to 12.0%63.9%, 6.7%63.6%, and 3.9%62.3%, in FE, LB, and AR, respectively; however, the ROM significantly increased compared with SPF in FE. Endpoint fixation with a cross-connector decreased ROM by 8.6%63.4%, 5.7%63.7%, and 3.2%62.4% in FE, LB, and AR, respectively, but was statistically equivalent to both SPF and EPF constructs. Fig. 2 compares the ROM for fixation methods. The CC translations were also significantly less than the intact and injured constructs. No statistical differences between the instrumentation methods were found for CC translation, unlike ROM. Segmental posterior fixation reduced CC translation to 5.1% 63.4%, 1.9%61.0%, and 1.5%60.6% during FE, LB, and AR, respectively. Endpoint fixation resulted in 11.0% 64.0%, 4.4%61.8%, and 3.8%62.3% during FE, LB,

and AR, respectively. The addition of a cross-connector for the EPFC construct changed CC translation to 9.9% 65.6%, 3.5%63.3%, and 3.0%61.8% during FE, LB, and AR, respectively. The trends of C0–C1 and C1–C2 during these rotations and translations slightly differ from C0–C2, especially when incorporating coupled motion. Since craniocervical injuries generally involve both joints, the data here was combined; however, individual joint motion can be seen in the Appendix. The intact and injured constructs are presented in Tables 1 to 3 of the Appendix, whereas coupled ROMs for the fixation groups during applied FE, LB, and AR are compared in Figs. 1, 3, and 5, respectively, of the Appendix. Translations of each joint for the fixation groups in anteroposterior, CC, and medial-lateral are also compared in the Appendix during FE, LB, and AR in Figs. 2, 4, and 6, respectively. Discussion Results from the present biomechanical study indicate that all fixation constructs significantly reduced motion compared with the intact and injury states. The injury model increased motion in three dimensions; however AR did

Table Intact and injury range of motion (degrees) of C0–C2 during applied moments in FE, LB, and AR Injured

Intact

C0–C2

FE  (SD)

LB  (SD)

AR  (SD)

FE  (SD)

LB  (SD)

AR  (SD)

25.4 (5.7)

8.4 (5.1)

71.8 (17.8)

42.8 (11.1)

22.2 (12.3)

93.5 (23.4)

FE, flexion-extension; LB, lateral bending; AR, axial rotation; SD, standard deviation. Significance (p!.05) is identified with bold.

1074

K.E. Radcliff et al. / The Spine Journal 15 (2015) 1070–1076

Fig. 2. Craniocervical junction range of motion with three different instrumentation methods. Significance (p!.05), if any, is shown as x, which is significant to intact and injured. Significance to SPF is shown as U. SPF, segmental posterior fixation; EPF, endpoint fixation; EPFC, endpoint fixation with a crossconnector; FE, flexion-extension; LB, lateral bending; AR, axial rotation.

not increase significantly. As expected, SPF with C1 lateral mass screws and C2 pars screws resulted in the greatest stability by significantly reducing motion. Instrumentation at every level between the occiput and C2 significantly reduced motion compared with EPF, where C1 is skipped, in FE. The addition of a cross-connector to the spanning construct improved not only the AR stability but also the FE. A previous biomechanical study showed that cross-connectors only add stability in AR; however, cross-connectors may have other benefits in cases of severe instability, such as shown in this work [17]. Cranial-caudal translations were also reduced significantly from the intact and injury states with all fixation constructs. However, unlike FE motion, there were no statistical differences between instrumentation methods’ CC translations. In the absence of screw failures, which would cause movement as the screw loosened from its boney purchase at the most outer ends, CC translations should be limited by the outermost fixation points at C0 and C2. In this study, the authors used a unique method to create the injury model. Prior models of CCJ instability have not created the injury from an axial approach nor included sectioning of the OC joint capsules. Some studies have used dens osteotomy models to mimic the injury created [6,24,28]. Two prior studies evaluated the instability after creation of a Type II odontoid fracture along with sectioning of the tectorial membrane, apical, and alar ligaments [19,29]. In contrast, this study included sectioning of the

occipitoatlantal joint capsules, which was based on a recent clinical study that found that occipitoatlantal joint capsule integrity predicts the stability of the occipitoatlantal joints [9]. One prior biomechanical study used an injury model that included sectioning of the occipitoatlantal joint capsules and other stabilizing ligaments, but anterior support structures were sacrificed to gain access to the internal ligaments [22]. The injury model in this study (sectioning the joint capsules, transverse ligament, alar ligaments, and tectorial membrane) resulted in increased rotation and translation at both joints, most significantly during LB and FE. Combined FE and AR forces enabled the occipital condyles to ‘‘clear’’ the C1 lateral masses and subluxate, which is why LB was the direction of greatest instability. The lateralized location of the joints at C0–C1 and C1–C2 naturally restrict LB motion. However, a ligamentous injury at these joints can cause separation to occur, resulting in significant increases in CC translations during applied LB. Prior studies have demonstrated that segmental fixation is the most rigid construct in the OC junction [6,19,22,24,28,29]. Uribe et al. [29] tested occipital condyle screws versus a standard occipital plate with C1 lateral mass and C2 pars screws in a cadaveric injury model. In contrast, this study questioned whether C1 segmental fixation was necessary and identified that a cross-connector may be an alternative solution when C1 fixation cannot be implemented. Studies have evaluated C1–C2 transarticular screws [6,22,24]. However, transarticular screws are

K.E. Radcliff et al. / The Spine Journal 15 (2015) 1070–1076

technically difficult to place in condition of OC instability, especially because of anterior translation of the C1 lateral mass relative to the C2 superior articular process, which may increase the risk of vertebral artery injury [30]. Therefore, although segmental instrumentation results in the most rigid construct, there is an inherent risk of neurovascular structure injury with more instrumentation. In patients with CCJ dislocation injuries, segmental fixation may be difficult or impossible because of patient anatomy. The most common scenario preventing C1 instrumentation during a segmental instrumentation procedure in CCJ injuries is vertebral artery injury or C1 fractures. Additionally, C1 instrumentation requires dissection and mobilization of the C2 venous plexus, resulting in a high risk of bleeding and necessitating additional operative time. Our data indicate that in CCJ dislocation injuries, occiput–C2 fusion without C1 fixation resulted in significantly reduced stability versus segmental fixation, which creates a challenge if the severity of the injury precludes C1 fixation. However, the addition of a cross-connector increased the rigidity of the occiput–C2 construct when skipping C1, making it statistically equivalent to segmental fixation. However, this was also not rigid enough to be statistically more rigid than EPF. Prior studies have demonstrated that two cross-connectors are the optimal stability construct in the spine [31–33]. Additionally, a prior thoracolumbar biomechanical study has demonstrated that cross-links increase stability of nonsegmental fixation to that equivalent to segmental fixation [31]. Collectively, these results indicate that in cases where C1 instrumentation is not possible or technically dangerous will reduce the biomechanical effectiveness of the fixation. In these cases, the utilization of a cross-connector will increase rigidity of the construct. The importance of segmental fixation is amplified when considering the stability of each individual joint. As the figures in the Appendix would suggest, removing C1 fixation causes significant increases in C0–C1 during EPF, which does not change by adding a cross-connector. Similarly C1–C2 is significantly less rigid with EPF than segmental fixation in secondary motion; however, unlike C0–C1, the cross-connector causes C1–C2 to be as rigid as segmental fixation. Other effects of coupled motion during applied bending motions are also affected by reducing to EPF (Appendix figures). Translations of all planes have minimal changes when comparing the fixations types, although some statistical differences were found (Appendix). Limitations of this study include the assumptions with the current biomechanical injury model. Muscles were excluded from all specimens and therefore, motions were applied using motors that may alter the kinematics of the spine. In vivo, passive muscle tone would function as a dynamic spine stabilizer. Additionally, there were other constructs, such as C2 translaminar screws or C2 pedicle screws, which could have been attempted. These constructs where chosen because C2 pars screws pose a low risk of vertebral artery injury with screw lengths less than 14

1075

mm. The injury model did not incorporate a fracture, which could have artificially added some stability in comparison with that seen in a full dislocation injury. The results, however, do provide a quantitative method to show which instrumentation would add maximum stability when severe instability is created at the CCJ. Conclusion Internal decapitation of the CCJ is a devastating, if not fatal injury. Stabilization of this injury proves to be a challenging surgical feat. Although segmental instrumentation can provide stabilization, including bilateral lateral mass screws at C1 vertebral body is not possible in all cases. This study assessed whether C1 lateral mass screws could be skipped without affecting the rigidity of posterior instrumentation. Overall, the strongest construct for stabilization of a craniocervical dislocation includes segmental instrumentation at the occiput, C1, and C2. Skipping instrumentation at C1 significantly increased motion in FE and resulted in more motion during LB. The addition of a cross-connector improved stability, such that it was statistically equivalent to segmental fixation. Based on these results, surgeons may balance the need for biomechanical rigidity and the safety of overall instrumentation constructs.

Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.spinee.2015.02.002 References [1] Tubbs RS, Hallock JD, Radcliff V, Naftel RP, Mortazavi M, Shoja MM, et al. Ligaments of the craniocervical junction. J Neurosurg Spine 2011;14:697–709. [2] Carlo Bellabarba SKM, Jens R. Chapman. Rockwood & Green’s fractures in adults. 6th Edition. Philadelphia, PA: Lippincott Williams & Wilkins, 2006. [3] Dickman CA, Mamourian A, Sonntag VK, Drayer BP. Magnetic resonance imaging of the transverse atlantal ligament for the evaluation of atlantoaxial instability. J Neurosurg 1991;75:221–7. [4] Dvorak J, Schneider E, Saldinger P, Rahn B. Biomechanics of the craniocervical region: the alar and transverse ligaments. J Orthop Res 1988;6:452–61. [5] Jain VK. Atlantoaxial dislocation. Neurol India 2012;60:9–17. [6] Sutterlin CE III, Bianchi JR, Kunz DN, Zdeblick TA, Johnson WM, Rapoff AJ. Biomechanical evaluation of occipitocervical fixation devices. J Spinal Disord 2001;14:185–92. [7] Menezes AH. Decision making. Childs Nerv Syst 2008;24:1147–53. [8] Menezes AH, Traynelis VC. Anatomy and biomechanics of normal craniovertebral junction (a) and biomechanics of stabilization (b). Childs Nerv Syst 2008;24:1091–100. [9] Radcliff K, Kepler C, Reitman C, Harrop J, Vaccaro A. CT and MRIbased diagnosis of craniocervical dislocations: the role of the occipitoatlantal ligament. Clin Orthop Relat Res 2012;470:1602–13. Epub 2011 Oct 29. [10] Bellabarba C, Mirza SK, West GA, Mann FA, Dailey AT, Newell DW, et al. Diagnosis and treatment of craniocervical

1076

[11]

[12]

[13]

[14]

[15] [16] [17]

[18]

[19]

[20]

[21]

[22]

K.E. Radcliff et al. / The Spine Journal 15 (2015) 1070–1076 dislocation in a series of 17 consecutive survivors during an 8-year period. J Neurosurg Spine 2006;4:429–40. Dublin AB, Marks WM, Weinstock D, Newton TH. Traumatic dislocation of the atlanto-occipital articulation (AOA) with short-term survival. With a radiographic method of measuring the AOA. J Neurosurg 1980;52:541–6. Horn EM, Feiz-Erfan I, Lekovic GP, Dickman CA, Sonntag VK, Theodore N. Survivors of occipitoatlantal dislocation injuries: imaging and clinical correlates. J Neurosurg Spine 2007;6:113–20. Bambakidis NC, Feiz-Erfan I, Horn EM, Gonzalez LF, Baek S, Yuksel KZ, et al. Biomechanical comparison of occipitoatlantal screw fixation techniques. J Neurosurg Spine 2008;8:143–52. Brown T, Reitman CA, Nguyen L, Hipp JA. Intervertebral motion after incremental damage to the posterior structures of the cervical spine. Spine 2005;30:E503–8. Deutsch H, Haid RW Jr, Rodts GE Jr, Mummaneni PV. Occipitocervical fixation: long-term results. Spine 2005;30:530–5. Grob D. Transarticular screw fixation for atlanto-occipital dislocation. Spine 2001;26:703–7. Majid K, Gudipally M, Hussain M, Moldavsky M, Khalil S. The biomechanical effect of transverse connectors use in a pre- and postlaminectomy model of the posterior cervical spine: an in vitro cadaveric study. Spine 2011;36:E1694–701. Coyne TJ, Fehlings MG, Wallace MC, Bernstein M, Tator CH. C1C2 posterior cervical fusion: long-term evaluation of results and efficacy. Neurosurgery 1995;37:688–92; discussion 92-3. Nassos JT, Ghanayem AJ, Sasso RC, Tzermiadianos MN, Voronov LI, Havey RM, et al. Biomechanical evaluation of segmental occipitoatlantoaxial stabilization techniques. Spine 2009;34:2740–4. Panjabi M, Dvorak J, Crisco J III, Oda T, Hilibrand A, Grob D. Flexion, extension, and lateral bending of the upper cervical spine in response to alar ligament transections. J Spinal Disord 1991;4:157–67. Panjabi M, Dvorak J, Duranceau J, Yamamoto I, Gerber M, Rauschning W, et al. Three-dimensional movements of the upper cervical spine. Spine 1988;13:726–30. Takigawa T, Simon P, Espinoza Orias AA, Hong JT, Ito Y, Inoue N, et al. Biomechanical comparison of occiput-C1-C2 fixation

[23]

[24]

[25]

[26] [27] [28]

[29]

[30]

[31]

[32]

[33]

techniques: C0-C1 transarticular screw and direct occiput condyle screw. Spine 2012;37:E696–701. Hirano K, Matsuyama Y, Sakai Y, Katayama Y, Imagama S, Ito Z, et al. Surgical complications and management of occipitothoracic fusion for cervical destructive lesions in RA patients. J Spinal Disord Tech 2010;23:121–6. Oda I, Abumi K, Sell LC, Haggerty CJ, Cunningham BW, McAfee PC. Biomechanical evaluation of five different occipitoatlanto-axial fixation techniques. Spine 1999;24:2377–82. Feiz-Erfan I, Gonzalez LF, Dickman CA. Atlantooccipital transarticular screw fixation for the treatment of traumatic occipitoatlantal dislocation. Technical note. J Neurosurg Spine 2005;2: 381–5. Harms J, Melcher RP. Posterior C1-C2 fusion with polyaxial screw and rod fixation. Spine 2001;26:2467–71. Goel A, Laheri V. Plate and screw fixation for atlanto-axial subluxation. Acta Neurochir (Wien) 1994;129:47–53. Frush TJ, Fisher TJ, Ensminger SC, Truumees E, Demetropoulos CK. Biomechanical evaluation of parasagittal occipital plating: screw load sharing analysis. Spine 2009;34:877–84. Uribe JS, Ramos E, Youssef AS, Levine N, Turner AW, Johnson WM, et al. Craniocervical fixation with occipital condyle screws: biomechanical analysis of a novel technique. Spine 2010;35:931–8. Paramore CG, Dickman CA, Sonntag VK. The anatomical suitability of the C1-2 complex for transarticular screw fixation. J Neurosurg 1996;85:221–4. Hart R, Hettwer W, Liu Q, Prem S. Mechanical stiffness of segmental versus nonsegmental pedicle screw constructs: the effect of crosslinks. Spine 2006;31:E35–8. Brodke DS, Bachus KN, Mohr RA, Nguyen BK. Segmental pedicle screw fixation or cross-links in multilevel lumbar constructs. A biomechanical analysis. Spine J 2001;1:373–9. Kuklo TR, Dmitriev AE, Cardoso MJ, Lehman RA Jr, Erickson M, Gill NW. Biomechanical contribution of transverse connectors to segmental stability following long segment instrumentation with thoracic pedicle screws. Spine 2008;33:E482–7.